Banking Industry Article

The Changing Face of Check Fraud

The transition from paper to imaged checks exposes bank vulnerabilities to fraud, but also facilitates detection methods that may turn the tide.

BYKAREN EPPER HOFFMAN
BAI Banking Strategies
September 11, 2009

Use of paper checks may be decreasing, but unfortunately the same cannot be said for check fraud.

As check payments undergo the rapid transition from paper to image, experts say that fraudsters are seizing on weaknesses exposed when banks convert both their processing and fraud detection/mitigation systems. As the physical check disappears from the process almost entirely, so too have many of the fraud detection capabilities that banks have used for years, including observing a check’s watermark, looking for micro-printing in the signature line and reviewing the look and feel of the check stock.

“The security features that we use in paper checks to identify potentially fraudulent items are being circumvented when we move to image,” says Jodi Pratt, founder of Jodi Pratt & Associates an Aptos, Calif.-based payments industry consultancy.

Indeed, bankers and industry analysts say that savvy fraudsters or criminal rings may even be targeting check payments more aggressively, taking advantage of these changes to run kiting scams or make duplicate check payments that they can present at multiple locations.

But while the move from paper to image may create some gaps in the short term, many payments experts believe the long-term outlook is more promising. The shortened timeframe for processing imaged payments and the emerging (and rapidly embraced) technologies for spotting fraud in the electronic check world are already decreasing the incidence of bad items. Software that can automatically look for check duplications or for invalid signatures can be applied more consistently and to a broader array of checks than was possible in the manual paper-based environment.

“The conversion of paper to electronic is making the fraud problem worse,” Pratt says, “but the technology has the potential to make it better. When you add image, you have more definitive factors to work with.”

Checks Go Away, Fraud Does Not Follow

Payments industry prognosticators have been heralding the death of the paper check for decades. And, it seems, the forecast is finally coming true as fewer consumers whip out their checkbooks and more of those transactions are translated to an image file as early as the point of sale or deposit.

“The transition to image processing is largely complete and the percentage of banks that are not receiving an image will be down to the single digits by the end of the year,” says Paul Obermeyer, executive vice president for the payment services group at Dallas-based Comerica Inc.

However, like a parasite that just gets stronger as its host weakens, the fraud inherent in the check payments system is proving resilient. While check volumes are decreasing at about 4% a year, by most industry estimates, attempted check fraud more than doubled between 2003 and 2006, reaching an estimated $12.2 billion in 2006, according to the American Bankers Association Deposit Account Fraud Survey Report, which was released in November 2007. Actual dollars lost annually increased nearly 50% during that timeframe as well, reaching nearly $1 billion in 2006. The 2009 Payments Fraud and Control Survey released by the Association for Financial Professionals earlier this year found that retailers are losing as much as $15 billion each year due to bad checks, the losses for which are largely passed along to the banks.

Obermeyer blames part of the problem on the “loss of some security features upon which we relied in the paper world,” particularly watermarking and micro-printing (the line where people sign their names, which is made up of tiny words and letters). In addition, he says, checks in the image world cannot be reviewed for color, check stock or the general look and feel that often helps banks differentiate between a fake and the genuine article. “Images don’t have the same degree of physical richness.”

According to Vijay Balakrishnan, president of StratEx LLC, an Atlanta-based business strategy consulting firm, the payments industry is divided on the question of whether the move to image is enabling banks to safeguard against fraud better than before (through speed and technology tools) or whether it will, in the short term, exacerbate the problem of check fraud.

“There are those who assert that technology allows the propagation of fraud at the speed of light; the paper check, after all, was bound by the limitations of planes, trains, and automobiles,” he wrote in a recent blog. “With access to the right software, images can be altered with greater ease than paper items.”

In an interview, Balakrishnan says that check kiting in the electronic realm can take place much faster than previously and that fraudulently presenting duplicate checks at multiple points of presentment is also easier.

Pratt says that savvy fraudsters are taking advantage of the fact that banks are at an “in-between stage” in the shift from paper to image. Many banks are so focused on converting their systems to be image-capable that they have not yet adopted new fraud detection/mitigation technologies and techniques to support this new channel, she adds.

“It’s All Been Telescoped”

Part of the problem of managing check fraud, as banks transition to image, is also likely a big part of the solution – the shortened processing window.

Balakrishnan points out that the electronification of check items, which greatly speeds up how fast they can be exchanged and clear, also shortens the time in which a fraudster can commit a crime. This “time compression” puts greater pressure on banks to do their detective work earlier on, he says.

“In the paper world, check fraud investigation is a day-two process,” Balakrishnan says. “In the image world, where everything is compressed, it needs to be day-zero. It’s all been telescoped.” In fact, more banks should be focused on placing their fraud detection technologies at the point of capture, at the teller station or with merchants, he adds.

David Walker, president and CEO of ECCHO, says that one of the best things banks can do to ultimately decrease fraud is to increase their ability to process checks faster. “What we’re seeing is banks taking advantage of that extra speed, and putting in better controls.”

Indeed, Mitch Christensen, executive vice president for payment strategies at Wells Fargo & Co., claims that check fraud has not increased at all for the San Francisco-based bank, which has been nearly 100% image-based on out-going and receiving checks since the third quarter of 2008. Fraud from check kiting is already slowing, he says, and will continue to decrease because of the “accelerated collection” process of image.

Jim Pitts, managing executive of the payments standing committee for the Financial Services Technology Consortium (FSTC), argues that while many check fraud detection technologies have emerged in the electronic realm, there are still many traditional detection methods that can be applied as well – just in a more automated way. For example, he says, instead of employees inspecting a check’s coloration and font and looking for microprint, computer software can handle that job.

“We can automate the process and look for the same red flags that indicate fraud,” Pitts says. “All these functions that were done hit and miss are now done automatically.”

Jim Resmer, senior vice president for BB&T Corp. of Winston-Salem, N.C., agrees that “some of the technology that’s been developed is a dramatic improvement over the manual process. If you’re relying on human beings, and someone’s looking at items for eight hours, they can lose their effectiveness.”

Christensen attributes Wells Fargo’s ability to hold down check fraud losses in part to a check duplication detection system on image files, which the bank developed itself and embraced early in the image process.

Detecting Anomalies

Balakrishnan says that he’s seeing more banks begin to embrace the use of tools to detect check alterations, once the transaction becomes an image file, as well as evidence of transactional fraud. He points to the growing availability of fraud detection tools similar to what the credit card industry has long embraced. These include fraud filters to detect anomalies in a customer’s normal pattern of check writing behavior or spending, negative file repositories such as Early Warning Services LLC (which can be integrated, in some cases, with distributed capture systems) to determine if a checking account might have been compromised and check duplication detection systems like the one Wells Fargo developed.

The need for duplicate detection is much more acute as banks move to image, Balakrishnan says, and the emergence of substitute checks means that “you could have three incarnations of the same check floating around,” if a fraudster tries to present it at various locations, for example. Eventually, he says, banks will need to utilize a duplicate detection system that can detect these issues across different institutions.

For banks that do not want to depend on homegrown solutions to help them detect and mitigate imaged check fraud, more vendors are stepping up to the plate. Milwaukee-based Metavante Corp., for example, offers a Teller Alert system to help banks detect fraud at the point of presentment in a branch. The system can pull information from both positive pay information and fraud filter systems (which can graph the transaction’s item amount and serial frequency and pull from lists of accounts with elevated suspicious activity) to confirm the validity of the deposit or alert the teller if the records do not jibe.

Parascript is a Boulder, Colo.-based company that develops technology to automatically analyze imaged items for accuracy, including the courtesy amount or legal amount written on checks. Mike Fenton, vice president of sales and operations for Parascript, predicts that the use of analysis software from his company and others will ultimately enable banks to review more aspects of the check for legitimacy. The software will also enable banks to review a higher percentage of checks that come through their systems and therefore solve the growing problem of check fraud that is perpetrated for smaller dollar amounts, he says.

Pratt says that the power of these new technologies to automatically and quickly review the various features of the check image with the transaction information that is provided will allow banks to tell, with a much higher degree of certainty, which items are likely fraudulent. Banks can also lower the rate of false positives that are triggered by traditional fraud detection methods. Banks that are combining these fraud detection tools are seeing as much as a 70% reduction in their rate of false positives, Pratt adds.

Ms. Hoffman is a freelance writer based in Poulsbo, Wash.